J. Electroanal. Chem., 68 (1976) 313--335 313 © Elsevier Sequoia S.A., Lausanne -- Printed in The Netherlands

ONE- VS. TWO-ELECTRON OXIDATIONS OF TETRAARYLETHYLENES IN APROTIC SOLVENTS

JOHN PHELPS and ALLEN J. BARD

Department o f Chemistry, The University o f Texas at Austin, Austin, Texas 78 712 (U.S.A.)

(Received June 5th, 1975; in revised form August 25th, 1975)

ABSTRACT

The oxidation of dimethylaminophenyl substituted ethylenes and tetra-p-anisyl ethylene at a platinum electrode in methylene chloride and acetonitrile solutions was investigated by cyclic and rotating disk electrode voltammetry, potential step chronocoulometry, con- trolled potential coulometry and e.s.r, spectroscopy. The results show that the oxidations occur by either two one-electron steps or by a single two-electron step depending upon the structure of the molecule. Factors influencing the reaction path are discussed.

INTRODUCTION

A question of fundamental interest in the electrochemical behavior of or- ganic compounds involves the factors which contribute to stepwise single- electron transfers:

R "+ + e ~ R E ° (1)

R 2 + + e ~ R t E ° (2)

vs. a single two-electron step

R 2+ + 2 e ~ K E ° = ( E ° + E ° ) / 2 (3)

In the absence of secondary chemical reactions or slow heterogeneous kinetics, which cause the observed waves to be shifted from their reversible positions, the difference between E ° and E ° is governed by solvation, ion pairing and structural effects [1,2]. In preliminary communications [1,3] we showed that the oxidation of aryl-substituted ethylenes could occur by stepwise one-elec- tron transfers or an apparent single two-electron transfer depending upon sol- vent and compound. A fuller discussion of this work and a more complete description of the experimental results are given here.

There have been relatively few studies of the electrochemical oxidat ion of aryl-substituted ethylenes in nonaqueous solvents. Schmidt and Meinert [4] found that 1,2-dif luoro-l , l -diphenylethane was formed during the constant current, oxidation of 1,1-diphenylethylene in a silver fluoride + acetonitrile

314

(ACN) solution. The reaction was envisaged as a two-electron oxidation fol- lowed by nucleophilic addition to the resulting dication. Mango and Bonner [5] and later Eberson and Nyberg [6,7] demonstrated that the oxidation of 1,1-diphenylethylene and stilbene in sodium acetate + acetic acid media also resulted in an oxidative addition to the double bond. They assumed the reac- tion involved an acetate assisted two-electron oxidation of the olefin. In a similar investigation, Inove et al. [8,9] examined the oxidation of stilbene and styrene in methanol. Although the products obtained also suggested ox- idative addition to the double bond, the proposed mechanism assumed initial oxidation of methoxide ion to the methoxy free radical, followed by free radical addition to the double bond. O'Connor and Pearl [10] suggested that the oxidation of 3,4-dimethoxypropenyl benzene in ACN occurs by formation of the radical cation followed by coupling and cyclization. Parker and Eberson [11] have reported a similar coupling reaction of 4,4'-dimethoxystilbene in ACN; the isolated product was analogous to that found for 3,4-dimethoxy- propenyl benzene. Sch~ifer and Steckhan [12] utilized this oxidative dimer- ization reaction to prepare several phenyl-substituted butanes and butadienes. A somewhat analogous reaction has been reported by Stuart and Ohnesorge [13,14] who found that tetraphenylethylene and tetraanisylethylene (TAE) undergo an oxidative cyclization to the corresponding aryl-substituted phen- anthrene. Parker et al. [15] proposed, based on cyclic voltammetric and coulo- metric data, that TAE in ACN undergoes a direct two-electron oxidation to form the stable dication. We previously contested this mechanism [3].

Related to the research described in this paper are studies of various di- methylamino-substituted ethylenes. Geske and Kuwata [16] conducted a polarographic and electron spin resonance (e.s.r.) investigation of the oxida- tion of tetra(N,N-dimethylamino)ethylene in ACN and dimethylformamide (DMF). This ethylene was shown to undergo a stepwise oxidation to the cation radical then to the dication. They obtained a well-resolved e.s.r, spectrum of the cation radical and showed that the formation equilibrium constant for the reaction

R 2÷ + R ~ 2R ÷ (4)

was about 230. Hi]nig et al. [17] studied a number of amino-substituted ethyl- enes. These generally exhibited stepwise oxidations with the formation con- stants of the radical cations, determined by polarographic measurements, being 103 to 10 s. Fritsch et al. [18] showed that 1,l-N,N-(dimethylamino)- ethylene is electrolytically oxidized in ACN to a radical cation which rapidly dimerizes and deprotonates to a butadiene, a reaction similar to the previously mentioned coupling reactions. They later described the electrooxidation and e.s.r, of a number of N,N-dimethylamino-substituted ethylenes [19] and found evidence of stepwise one-electron and single two-electron oxidations. The chemical oxidation of tetraamino-substituted ethylenes and the properties of the resulting dications have been reviewed by Wiberg and Buchler [20] and Hoffman [2 l ].

315

EXPERIMENTAL

Chemicals

The solvents used in this work were acetonitrile (ACN) and methylene chloride (MC). Practical grade ACN, obtained from Matheson, Coleman and Bell, was purified by a modification of a previously described procedure [22]. The MC was Baker Chemical Company or Aldrich Chemical Corporation reagent grade. In general, the solvent could be used as received. However, an acidic impurity appeared in some batches; its removal was effected by a sim- ple fractional distillation from calcium hydride [23].

The supporting electrolyte used in most experiments was polarographic grade tetra-n-butylammonium perchlorate (TBAP) obtained from Southwest- ern Analytical ' Chemical Company. It was dried in a vacuum oven at 100°C for two days, then stored over anhydrous magnesium perchlorate. The other supporting electrolytes used were lithium perchlorate and sodium perchlorate obtained from G. Frederick Smith Chemical Company. These salts were puri- fied by multiple recrystallizations from water, followed by dehydration in a vacuum oven at 150°C and storage over anhydrous magnesium perchlorate.

N,N,N',N'-tetramethyl-p-phenylenediamine (TMPD) was prepared by neu- tralization of the dihydrochloride obtained from Eastman Organic Chemical and was sublimed before use. Tetra-p-anisylethylene {TAE) was kindly sup- plied by Dr. E.A. Chandross. Tetrakis(p-N,N-dimethylaminophenyl)ethylene (OMPE) was prepared by the method of Gatterman [24]. Cis- and trans-l ,2- bis(p-N,N-dimethylaminophenyl}-l,2-diphenylethylene (sTMPE) was prepared by the method of Buckles and Meinhardt [25]. Although these authors re- ported only one isomer from the reaction {coupling of p-N,N-dimethylamino- benzophenone with tin and HC1), we obtained both the cis and trans forms. The isomers were separated by fractional crystallization from 1:1 benzene/ Skelly B, yielding a more soluble cis-isomer (m.p. 201.5--205.5°C), and a less soluble trans-isomer (m.p. 222--230°C), which produced distinctly different n.m.r, spectra. 1,1-Bis(p-N,N-dimethylaminophenyl)-2,2-diphenylethylene (uTMPE) was prepared by a modification of the method of SchSnberg [26].

Preparation of the previously unreported 1,1,2-tris(p-N,N-dimethylamino- phenyl)-2-phenylethylene (HMPE) was as follows. The diazo compound pre- pared from 2 g {0.008 mole)p-dimethylaminobenzophenone hydrazone was added to a benzene solution of 2.4 g (0.008 mole) 4,4'-bis(dimethylamino)- thiobenzophenone, and the mixture was stirred for two days at room tem- perature. After the benzene solution was concentrated and chromatographed on an alumina column, eluting with benzene, 3.1 g {77%) of a tan colored episulfide was obtained. The desulfuration of 0.36 g (0.0007 mole) episulfide with 0.53 g (0.002 mole) triphenylphosphine in 25 ml p-xylene yielded the ethylene. Two recrystallizations from ethanol resulted in 0.23 g (70%) of the ethylene with m.p. 193--194.5°C; u.v. max {CHC13) 365 nm {1.82 X 104); n.m.r. (CDC18) , 2.8 ~i, two incompletely resolved singlets (18 H total), 6.5 multiplet (12 H) and 6.9 ~i, singlet (5 H); mass spectrum m / e = 461, m / 2 e =

316

230.5. Anal.: calcd, for C32H35N3 : C, 83.26; H, 7.64; N, 9.10. Found: C, 83.49; H, 7.52; N, 8.72.

In general all of the prepared compounds were purified by recrystallizations and chromatography; details of the preparations and characterizations are available [26].

Apparatus

Most voltammetric experiments were performed on an instrument similar to that described by Anson and Payne [ 27 ], equipped with positive feedback for /R-compensat ion. The compensation was adjusted as follows: a potential step was applied to the cell, the compensation increased until damped oscilla- tions were observed in the current--t ime curve, and then the compensation re- duced until a smooth curve was obtained. Controlled potential coulometric ex- periments utilized a Wenking-type potent iosta t equipped with a Krohn-Hite DCA-10 amplifier. The current--t ime curve was displayed on a Varian Model G-14 strip chart recorder; current integration was accomplished using the di- gital readout system described by Bard and Solon [28].

Vacuum cells and vacuum line preparation techniques [29] were employed for all experiments. The working electrode for most of the transient experi- ments was a platinum disk electrode of geometric area 0.013 c m 2. The elec- t rode was fabricated by sealing a plat inum wire in a 6 mm soft glass tube, then grinding both platinum and glass to a planar surface using carborundum grind- ing powders on a glass plate. Contact with the platinum wire was made with mercury and a copper wire. Small scratches left in the platinum by the grinding process were removed by polishing with a Superior H.R. water soluble red rouge. The e lect rode was then oxidized with nitric acid, reduced with acidic ferrous ammonium sulfate solution, and cycled in the supporting electrolyte solution. The working electrode for the coulometric experiments was a plati- num gauze electrode; this same gauze was used as the counter electrode in all other experiments. The counter electrode for coulometry was a plat inum spiral placed in an 8 mm fritted glass tube. The fritted counter electrode com- par tment usually was filled with an aqueous potassium nitrate agar salt bridge for additional separation. An aqueous saturated calomel electrode (SCE) with a 3% agar + 1 M potassium nitrate salt bridge terminated with a fine porosity frit was used as the reference electrode.

In e.s.r, examination of a coulometric solution, a sample was removed into an evacuated 3 mm Pyrex tube at tached to the electrochemical cell. The sam- ple was frozen in liquid nitrogen, and the tube was sealed and removed with a torch. After the sample thawed, the e.s.r, experiments were carried ou t on Varian Models V-4502 or E-4 spectrometers.

RESULTS

The tetraarylethylenes studied are shown in Fig. 1. Experiments were also conducted on a model compound whose electrochemical behavior has been

317

R1 0 R~ TMPD

~ C C "~R~

RI tPanS -sTMPE

RI~c C ~RI

cis -sTMPE

R1 H MPE

RI~C C . ~ R1

OMPE

R2'~C~c~R2

R ~ -- " ~ R2 TAE Fig. 1. C o m p o u n d s in this s tudy; R 1 = N(CH3)2, R 2 = OCH 3.

well characterized: N,N,N',N'-tetramethyl-p-phenylenediamine (TMPD). Dvo]~k et al. [30] have shown that the oxidation of TMPD in ACN occurs by a stepwise process involving two, well-separated, one-electron steps. The first step is a reversible oxidation of TMPD to a stable cation radical; this ox- idation will serve as a model for a one-electron oxidation in the fol lowing dis- cussions.

The aryl-substituted ethylenes can be classified as belonging to one of the following groups: (a) two well-separated one-electron processes, (b) two

318

poorly separated, unresolved, one-electron processes, or (c) direct two-electron processes, or at least two one-electron processes in which the second occurs more readily than the first, depending on their behavior during oxidation.

Cyclic and rotating disk electrode voltammetry in methylene chloride

The cyclic voltammetric behavior of the compounds of interest in MC solu- tion is summarized in Table 1. Representative voltammograms are given in Fig. 2 and detailed tables of data at varying scan rates are given in ref. 26. Of the tetraarylethylenes considered, only uTMPE shows two well-separated waves, each corresponding (by comparison to TMPD) to one-electron oxida- tions. The current function for the first wave is slightly smaller than that for TMPD because of the smaller diffusion coefficient of uTMPE. The second ox- idation wave for uTMPE is irreversible. The current function for the second wave, when measured from the decaying current of the first wave (Fig. 2a) is about 10% larger than that of the first wave. The peak width, Epa--Epa/2 , is 50 mV. These parameters correspond more closely to a reversible one-electron charge transfer followed by a fast chemical reaction than to those expected for a slow charge transfer step. Thus uTMPE belongs to group (a) and involves a reversible oxidation to a fairly stable radical cation at the first wave and reversible oxidation to a dication which undergoes a rapid decomposition reaction at the second wave.

The cyclic vol tammetry of TAE (Fig. 2b) clearly demonstrates the presence of two closely spaced waves. Because of the poor separation of the two waves, the data in Table 1 reflect only the peak currents and peak potentials of the composite wave. The current funct ion of 16.9 for this composite wave, com- pared to the one-electron value of about 13 and the two-electron value of 31 to 34 (see below) for these compounds, is consistent with a mechanism in- volving two closely spaced one-electron waves [31] [group (b) behavior]. A comparison of the wave with theoretical ones for stepwise electron transfer yields a mechanism in which TAE is oxidized to a stable radical cation and then to a stable dication, with the E ° of the second step about 90 mV more positive than that of the first.

The other substituted ethylenes, sTMPE, OMPE, and HMPE, which involve at least one dimethylaminophenyl group on both ethylenic carbons, all show very similar behavior and a single oxidation wave (Fig. 2c). The peak potential separations, Epa--Epc , are 30 to 40 mV and the current function is about (2) a/2 times that of uTMPE. These results are consistent with single reversible two- electron oxidations to stable dications, i.e., group (c) behavior. Although the results at the 1 mM concentration level show reasonably constant current functions and ipc/ipa ratios with scan rate, at lower concentrations a distinct decrease in current function and increase in ipe]ipa with scan rate is observed (Table 2). These results are best explained by weak adsorption of the dication, since the cyclic voltammetric curves generally agree with the theoretical treat- ment of this mechanism by Shain and Wopschall [32] and Laviron [33]. This

TA

BL

E 1

Su

mm

ary

of

cycl

ic v

olt

amm

etri

c an

d r

ota

tin

g d

isk

elec

tro

de

dat

a a

Compound b Concn./mM

C.v

RD

E f

i v-

-1/2

c--l

! "

E1/

2/V

L

og

plo

t e

iL~O

--1/

2c--

1 f/

E

p a/

V

Ep

c ]V

~

P/~

I/2V

--1/

~m

M--

1 lP-

--~c

(vs.

SC

E)

(vs.

SC

E)

/pa

(vs.

SC

E)

slo

pe/

mV

pA

s--1

/2

mM

--1

Gro

up

TM

PD

3

.69

0

.21

0

.14

14

.9

1.10

a

uT

MP

E c

1.

71

0.6

1

0.5

5

12

.9

1.0

2

0.5

68

58

0

.87

a

TA

E

1.8

4

0.9

9

0.8

5

16

.9

1.0

4

0.9

11

g

1.52

b

sTM

PE

d

1.6

5

0.4

4

0.41

3

3.8

1

.05

0

.44

0

34

2.0

6

c O

MP

E

1.0

3

0.1

8

0.1

4

31.1

1

.02

0

.15

8

36

1.81

c

HM

PE

1

.32

0

.32

0

.28

3

1.0

1

.02

0

.32

1

36

1.94

c

a T

he

solu

tio

n w

as 0

.2 M

TB

AP

in

met

hy

len

e ch

lori

de

and

th

e w

ork

ing

ele

ctro

de

was

a P

t di

sk,

geo

met

ric

area

0.0

13

cm

2.

Ave

rage

re-

su

lts

for

scan

rat

es,

v, b

etw

een

0.0

5 a

nd

1 V

s -1

. b

See

Fig

. 1.

c

Fo

r se

con

d w

ave,

Ep

a =

0.8

8 V

vs.

SC

E,

ipa

(2)/

v112

c =

14 a

nd

ipa

(2)/

ipa(

1 )

= 1.

1. E

l~ 2

= 0

.85

Vvs

. S

CE

, lo

g p

lot

slop

e -1

=

50 m

V

and

iL/

¢01/

2c

= 1.

64.

d C

/s a

nd

tra

ns f

orm

s sh

ow

vir

tual

ly i

den

tica

l el

ectr

och

emic

al p

rop

erti

es.

e R

ecip

roca

l o

f sl

op

e o

f lo

g (i

/i L

--

i) v

s. E

. f

Ro

tati

on

rat

e, c

o, 1

00

0 r

pm

. g

No

nli

nea

r, s

ee t

ext.

~D

320

. u

Z LLI 0:: 0:: :::)

12

4 ,< :3.

o I - Z w - 4 0¢ e,-

O _ 6

._u

k-- Z t~J

n - .-~

(.3

a

i.o 0-8 0-6 0.4 0.2 E/V. vs SCE

i

ll'O OES/ V. 01"6vs. scEO~4 0~2

T 4jJA

l

o.7 0'.6 o'.5 6

c

(

o~4 o!3 . o~2 o!t E/V vs SCE

Fig. 2. Cyclic voltammetric oxidat ion of (a) uTMPE, (b) TAE and (c) HMPE in 0.2 M TBAP + methylene chloride at plat inum electrode.

TABLE 2

Cyclic voltammetric data for 1,1,2-tris(p-N,N-dimethylaminophenyl)-2-phenylethylene (HMPE) a

Scan Epa/V Epc/V ipa/PA ipc/PA ipc/ipa pA sl/2V--1/2m_M -1 rate/V s - 1 (vs. (vs. ipaV--1/2c--1/

SCE) SCE)

(A). Oxidation o f HMPE Concentration: 1.32 mM 0.051 0.31 0.275 9.4 9.6 1.02 31.5 0.102 0.31 0.275 13.3 13.4 1.01 31.8 0.152 0.31 0.275 16.2 16.5 1.02 31.6 0.203 0.31 0.275 18.9 19.5 1.03 31.8 0.254 0.31 0.27 21.2 21.5 1.01 31.8 0.304 0.32 0.275 22 22.5 1.02 31 0.406 0.32 0.27 25 26 1.04 30 0.508 0.32 0.27 28 28.5 1.02 30 0.762 0.32 0.27 34 35 1.01 30 1.016 0.32 0.27 39 40 1.01 30 Concentration: 0.41 mM 0.051 0.305 0.28 2.82 3.0 1.1 30.4 0.102 0.305 0.28 4.00 4.4 1.1 30.8 0.152 0.31 0.275 5.02 5.5 1.1 30.4 0.203 0.31 0.275 5.50 6.0 1.1 29.8 0.254 0.31 0.27 6.05 6.7 1.1 29.4 0.304 0.315 0.27 6.65 7.3 1.1 29.5 0.356 0.32 0.26 6.9 7.8 1.1 29 0.406 0.32 0.26 7.4 8.4 1.1 29 0.457 0.32 0.26 7.8 8.9 1.1 28 0.508 0.32 0.26 8.2 9.6 1.2 28 0.762 0.32 0.26 10.0 12.0 1.2 28 Concentration: 0.10 mM 0.051 0.325 0.29 0.61 0.64 1.1 27 0.102 0.325 0.29 0.86 0.93 1.1 27 0.152 0.325 0.285 1.04 1.16 1.11 27 0.203 0.325 0.285 1.20 1.40 1.17 27 0.254 0.325 0.29 1.35 1.55 1.15 27 0.304 0.325 0.285 1.47 1.75 1.19 27 0.356 0.325 0.28 1.52 1.19 1.25 25 0.406 0.325 0.28 1.64 2.0 1.2 26 0.457 0.325 0.28 1.74 2.2 1.3 26 0.508 0.325 0.28 1.84 2.3 1.3 26 0.762 0.325 0.28 2.18 3.0 1.4 25 1.016 0.325 0.275 2.44 3.55 1.5 25

(B) Reduction o f HMPE solution totally oxidized at +0.6 V vs. SCE ipa/ipc ipcv--ll2c--1/I2A sl/2V--1/2mM--1

Concentration: 0.97 mM 0.051 0.340 0.300 5.36 5.40 0.99 24.6 0.102 0.340 0.300 7.52 7.68 0.98 25.0 0.152 0.340 0.300 9.20 9.36 0.98 24.8 0.203 0.340 0.300 10.7 10.9 0.98 24.8 0.254 0.34 0.30 13.2 13.7 0.97 25.6 0.304 0.34 0.30 14.1 14.2 0.99 24.4 0.406 0.34 0.30 15.0 15.2 0.98 24.6

a The solution was 0.2 M TBAP in CH2CI 2. The working electrode was platinum disk with geometric area 0.013 cm 2.

322

I 1.2 1.0 0.8 0,6 0.4

E/V vs SCE

T 4pA

1

1.0 b

0.5

~'- 0

-0.5

-1.0 o'~ o o~ o

0.2g 0.30 0.31 0.2 0.33 0.34 0.35 i i | l i

/

~/JA 1

I 0.8 ~6 0.4

E/V v$ SCE

Fig. 3. Rotating disk voltammetry of (a) uTMPE and (b) HMPE in 0.2 M TBAP + methylene chloride. Also shown in (b) is logarithmic analysis of HMPE wave.

adsorption was studied in more detail by double potential step chronocoulo- merry.

Typical voltammograms at the rotating disk electrode (RDE) are shown in Fig. 3; the data are summarized in Table 1. The results generally confirm those of cyclic vol tammetry; the deviation of the slope for the group (c) compounds from the expected theoretical value of 29 mV probably reflects some uncom- pensated/R-drop. The log plot for TAE was curved, as expected for a voltam- mogram of two unresolved waves. The experimental points could be fit very well (Fig. 4), using the polarographic t reatment of Miiller [34] and Brdi~ka [35] assuming two reversible electron transfers with an equilibrium constant of the reaction

TAE 2+ + TAE ~ 2TAE'* (5)

of 55, with E ° 's for the first and second steps of 0.86 and 0.96 V vs. SCE, respectively.

O.e o/ o/O/°/

o / o /

I | 0.85 0.90

E/V vs SCE

Fig. 4. L o g a r i t h m i c analysis o f the e u G e n t - - p o t e n t i a l curve resu l t i ng f r o m the o x i d a t i o n o f TAE in methylene chloride at a platinum RDE. The points represent experimental data and the line represents the theoretical curve for K = 55.

.~ o.c

- J

-0.~

TA

BL

E 3

Po

ten

tial

ste

p c

hro

no

amp

ero

met

ry

and

ch

ron

oco

ulo

met

ry

dat

a su

mm

ary

a, b

Co

mp

ou

nd

E

I/V

E

2/V

it

-1/2

slo

pe/

nK

C/p

A

K e

lizA

(v

s. S

CE

) (v

s. S

CE

) /.t

A s

1/2

s 1

12 m

M -1

s 1

12 m

M -1

Q

t 1/2

slo

pe/

pC

s -1

/2

nK' c

~ K

' c~

pC s

-1/2

pC

s -1

/2

mM

--1

raM

--1

K'/

K d

TM

PD

3

.69

mM

--

-Q.1

0.

5 u

TM

PE

1

.71

mM

0

.0

0.7

4.0

6 e

2

.38

2.

38

0.0

1.0

8.0

0

4.6

8

2.3

4

OM

PE

1

.03

mM

--

0.2

0

.6

4.7

4

4.6

0

2.3

0

HM

PE

0

.97

mM

0.

0 0.

6 tr

ans-

sTM

PE

1

.59

mM

0

.0

0.8

7.2

5

4.5

6

2.2

8

c/s-

sTM

PE

2

.18

mM

0.

0 0.

7 9

.43

4

.33

2

.11

21

.8

7.6

5 f

15

.6

8.1

2

18.1

5.9

0

5.9

0

4.4

8

4.4

8

9.1

0

9.1

0

8.3

6

8.3

0

4.1

8

4.1

5

1.8

8

1.9

5

1.9

7

. T

he

solu

tio

n w

as 0

.2 M

TB

AP

in

CH

2CI 2

. b

Th

e w

ork

ing

ele

ctro

de

was

a p

lati

nu

m d

isk

of

geo

met

ric

area

0.0

13

cm

2.

c T

hes

e p

aram

eter

s ar

e d

efin

ed a

s~K

= F

AD

ll2/

?rl/

2n a

nd K

' =

2F

AD

1/2/

~I/

2n.

d T

his

par

amet

er h

as a

th

eore

tica

l v

alu

e o

f 2.

e

Rat

io o

f sl

op

es =

1.9

7.

f R

atio

of

slop

es =

2.0

3.

b~

c~

324

Potent ia l s tep c hronoamperome t ry and chron o co u lo m e t ry in m e th y l en e chloride

Potential step measurements were undertaken to obtain still more con- firmatory evidence about the reaction steps, to probe for adsorption of reac- tants and products, and especially to determine the effectiveness of resistance compensation in MC. The results were evaluated from i v s . t -112 and Q v s . t 112

plots, and the electrode diffusion current constants K and K', defined in (6) and (7) were determined. Typical experimental results are given in Table 3 and Fig. 5. The constancy of the

it~/2/CR = FAD1/2/Tr 1/2 = n K (6)

(Q _ Qdl)/ t l /2 C R = 2FAD1/2/Trl/2 = nK ' (7)

K and K ' values for the different compounds with the n-value assumed from the voltammetric studies confirms these values and demonstrates that these compounds have similar diffusion coefficients. Note also that the ratio K ' / K is close to the theoretically expected value of 2. For uTMPE the ratio of slopes for steps to potentials on the second wave (1.0 V) compared to those at the first (0.7 V) were also near the value of 2 expected for stepwise reactions in- volving an equal number of electrons. The i - t - 1 / 2 and Q - - t 1/2 plots were linear for times between 10 ms and 2 s. At shorter times a positive deviation was observed, probably because of the presence of uncompensated resistance which prevents rapid double layer charging. The positive deviation at times

m .I .2 f

0 1 2 T~'2/S 1/2

Fig. 5. C h r o n o c o u l o m e t r i c charge vs. t ime1/2 p]o t for the o x i d a t i o n o f c i s - s T M P E in me thy l - ene ch lor ide at a p l a t i n u m e lec t rode . T he inser t r ep resen t s an e x p a n s i o n of the p lo t for t ime less t h a n 100 ms.

u 30 &

o

dr

longer than 2 seconds probably arises from convection due to vibrations of the cell while connected to the vacuum line.

A double potential step chronocoulometric experiment with HMPE was also performed to estimate the extent of adsorption of the dication. A 0.25 mM solution of HMPE dication was prepared by controlled potential oxida- tion of the parent at +0.6 V. The results of a double potential step (0.5 to 0.0 to 0.5 V) are shown in Fig. 6. The results were analyzed [36,37] assuming t h a t the dication was adsorbed but the product (parent) was not. Under these conditions the charge during the forward (0.5 to 0.0 V) step is given by

Q~ = Q(t < T) = nK'co tl/2 + n F A F o + Qdl (8) where Co is the bulk concentration of dication and Fo the amount of dication adsorbed. At' t ime t = I" the potential is stepped back to 0.5 V and the charge on reversal is given by

Qb = Q(r) -- Q( t > r) = nK'co (1 + n F A a l F o / Q c ) 0 + a o n F A F o + Qdl (9)

where 0 = (t -- T) z/2 + T 1 / 2 - - t 1 / 2 and a0 and al are the slope and intercept of the plot of 1 -- (2Dr) sin -1 [(r / t ) 1]2 ] vs. O/T 112 [36]. The plot of Fig. 6 yields the following results

Slope Q~ -- t 1/2 - Sf = 2.02 pC s -z/2

Q~ intercept _- OQ~ = 0.122 pC

--,,.,,, &

cf

0.2

0.8

0 .6

0.4

0.2

<o T~/2/S 1/2 0.1 0.2 0.3

i i i

e

0.4

325

Fig. 6. Charge vs. time1/2 and charge vs. 0 plots for double potential step chronocoulometry for a solution of HMPE dication in methylene chloride. The potential was stepped from 0.5 to 0.0 to 0.5 V vs. SCE.

326

Slope Qb -- 0 -- SD = 2.22 pC s -1/2

Qb intercept -- °Q b = 0.055 pC

Qc (Cottrell charge) ~ 2K 'T 112 = 0.64 pC

From these parameters, the amount of adsorbed dication calculated by eqn. (9) was 2.5 × 10 -11 moles cm - 2 . This amount adsorbed represents less than monolayer coverage. Using the internal consistency check advocated by Chris- tie et al. [36], the measured slope ratio Sb /S f was 1.10 which agrees exactly with the theoretical value.

Qdl was found to be 0.06 pC which was somewhat less than the 0.08 pC double layer charging term measured in a blank solution under identical con- ditions.

T A B L E 4

Cou lome t r i c da ta o n se lec ted t e t r a a r y l e t h y l e n e s in m e t h y l e n e ch lor ide a

C o m p o u n d O x i d a t i o n nap p con t r o l p o t e n t i a l / V (vs. SCE)

Reversal c o u l o m e t r y b

R e d u c t i o n nap p con t ro l p o t e n t i a l / V (vs. SCE)

OMPE 1.16 m M 0.6 2.01 - -0 .4 2.06

HMPE 1.51 m M 0.8 2 .04 0.0 1.96 c 0.97 m M 0.6 2.00 - -0 .2 1 .99

t rans-sTMPE 1.53 m M 0.6 2.01 0.0 1.98

cis-sTMPE 1.58 mM 0.8 1.98 0.0 2 .02

uTMPE 1.57 m M 0.7 1.1 d 0.0 0.1 3 .36 m M 0.7 0 .98

TAE 0.44 m M 0.95 0 .86 e

1 .20 2.1

a The so lu t i on was 0.2 M TBAP. b These da ta refer to reversal c o u l o m e t r y on the c o m p l e t e l y ox id ized paren t . In general ,

t h e t ime b e t w e e n c o m p l e t i o n of o x i d a t i o n and in i t i a t ion o f r e d u c t i o n was a b o u t one hour . c The ox id i zed so lu t i on was a l lowed to r e m a i n overn igh t ; t he r e d u c t i o n was in i t i a t ed the

nex t morn ing . d Elec t ro lys is was t e r m i n a t e d a t a s t eady-s ta te c u r r e n t of a b o u t one p e r c e n t of t he ini t ial

cur ren t . e E.s.r. e x a m i n a t i o n ind ica ted the presence o f an in tense s ingle t radical spec t rum, w i t h some

s u p e r i m p o s e d h y p e r f i n e s t ruc ture .

327

Controlled potential coulometry in methylene chloride

Coulometry experiments at a platinum gauze electrode controlled at a po- tential on the limiting current plateau were undertaken to determine definitive nap p-values, to determine if any slow chemical reactions of the primary elec- trode reaction product occurred, and to prepare solutions of the electrode reac- tion product for further vol tammetry studies. The results are given in Table 4. The type (c) compounds all show napp-values near 2.0 corresponding to the oxidation of the parent to the intensely colored dication. As expected, no e.s.r, spectrum was observed for a completely oxidized solution of OMPE. To determine the long-time stability of the type (c) dications, reversal coulometry [38] experiments were conducted on the completely oxidized solutions. In all cases, napp-Values near 2.0 were found. In one experiment, a 1.51 mM HMPE solution was completely oxidized, then left overnight exposed to the atmo- sphere. Twelve hours later, reversal coulometry showed n a p p = 2.0. These ex- periments provide conclusive evidence that the group (c) compounds are reversibly oxidized to extremely stable dications. Voltammetric studies on the oxidized solutions will be given below.

The oxidations of the type (a) and type (b) compounds were quite different. Oxidation of TAE at 1.2 V vs. SCE gave an napp-Value of 2.1. However oxida- tion at 0.95 V gave an n~pp-value of 0.86; e.s.r, examination of this solution showed an intense singlet with some superimposed fine structure. This cation radical has also been obtained by chemical oxidation with methanesulfonic acid in methylene chloride, and its interpretation has been reported [39].

The controlled potential oxidation of uTMPE at 0.7 V vs. SCE showed an n~pp of about one. An e.s.r, examination of the oxidized solution showed the presence of a radical, but the signal was very weak and not well resolved. Ap- parently, the radical decomposes in a chemical reaction that is slow compared to the voltammetric time scale, i.e., order of seconds, but that is important compared to the coulometric time scale, i.e., order of for ty to sixty minutes. Reversal coulometry on this oxidized solution showed only one tenth of an electron per molecule was recovered, indicating extensive decomposit ion of the radical cation.

Voltammetry of oxidized solutions in methylene chloride

An RDE voltammogram of an 0.97 mM HMPE solution completely oxidized at 0.6 V vs. SCE showed a cathodic wave with an E1/2 of 0.321 V vs. SCE, a Levich constant of 1.63 and a log plot slope -1 of 38 mV. These data can be compared with data for the original solution which showed an oxidation wave with a n El~ 2 of 0.318 V vs. SCE, a Levich constant of 1.94 and log plot slope -1 of 36 mV. With the exception of a lower limiting current, these data are con- sistent with the reversible two-electron reduction of the dication to the orig- inal compound. Similarly the cyclic voltammogram of the totally oxidized HMPE solution was essentially the same as that for the parent solution except

3 2 8

that the peak currents were about 11% smaller for the oxidized solution (Table 2B). Note also that the ipc/ipa ratio is now slightly less than 1.0, since it is the reactant which is now weakly adsorbed.

The difference in limiting and peak currents between parent and dication suggests that either the concentration of the dication is lower than that of the original parent or that the diffusion coefficient of the dication is considerably smaller than that of the parent. Loss of dication by precipitation of the salt on the large gauze electrode is a possibility; no precipitate was observed in solution or on the electrode surface, however. Note that reversal coulometry showed essentially complete recovery of the dication. The amount of adsorp- tion of the dication determined by the chronocoulometric experiment is much too small, even with electrode areas of 300 c m 2 to account for any appreci- able decrease in the solution concentration. A difference of diffusion coef- ficient of about 25% between parent and dication is required to account for the limiting current differences. While this seems unusually large, it may be caused by the dication existing in the form of an ion pair or ion aggregate with perchlorate ion in this low dielectric constant solvent.

Because of the difficulty of making quantitative correlations of the com- posite waves with a theoretical model, extensive vol tammetry studies on ox- idized solutions of TAE were not conducted. Cyclic vol tammetry of the com- pletely oxidized solutions demonstrated the presence of a reasonably stable dication which was reduced in the same stepwise manner as the parent was ox- idized. Cyclic voltammetric reduction of an oxidized solution of uTMPE showed only a small wave for the reduction of uTMPE cation radical. This observation reflects the extensive decomposition of the cation radical already suggested by the reversal coulometry.

Direct product isolation from the oxidized solutions was unsuccessful as all at tempts to separate the dication salt from the supporting electrolyte failed. However, similar products have been isolated from the chemical oxidation of these compounds. Baenziger et al. [40] have reported the isolation of stable dication salts of TAE. Anderson et al. [41] have reported the isolation of the dinitrate and diiodide of OMPE by oxidation of the parent with silver nitrate.

Although direct isolation proved difficult, some information conceming the structure of the dication can be obtained by reducing the dication back to the neutral parent, then isolating the starting material. Product isolation from the reduction of an oxidized solution of OMPE and of HMPE was effected by stripping the MC from the electrolysis solution, extracting the residue with carbon disulfide in a Soxhlet extractor to separate the products from the sup- porting electrolyte, and finally stripping the carbon disulfide. The product was identical to starting material by n.m.r, and i.r.; recovered yields were 70--90%.

The product of the oxidation--reduction of sTMPE was more revealing. A 2.14 mM solution of cis-sTMPE was subjected to complete oxidation and sub- sequent complete reduction, and the products were isolated. An n.m.r, spec- trum of the product dissolved in carbon disulfide demonstrated that a mixture of isomers was recovered, which was quite similar to that obtained from the

329

initial preparation of these substituted ethylenes. Apparently the dication isomerizes in solution. This observation suggests that the bond order of the central ethylene bond in the dication decreases considerably and thus allows free rotat ion about this bond.

O x i d a t i o n s in a c e t o n i t r i l e

In general, the voltammetric data for the oxidation of the type (c) com- pounds in MC deviated slightly from theoretical expectations for a direct two- electron oxidation. For example, the peak potential separations were about 10 mV greater than theoretical predictions and slopes -1 of log i / ( iL - - i) vs. E plots were greater than expected, i.e., 34--36 mV compared to 29.5 mV. These differences could be caused by uncompensated iR drops in the measured po- tential values and by the reactions occurring by two very closely spaced one- electron transfers, where E ° is equal to or slightly less than E ° . To investigate these possibilities e.s.r, experiments were conducted on partially oxidized OMPE solutions to search for radical cation; these are described below. In addition voltammetric studies were conducted on OMPE and TAE in acetonitrile (ACN), where uncompensated/R-effects are negligible with our apparatus, and where the effects of solvation on the nature of the electrode reactions could be probed. The cyclic vol tammetry of OMPE in ACN showed only a single, re- versible wave. The pertinent vol tammetry results (Table 5) are as follows: the

TABLE 5

Typical cyclic voltammetric data for the oxidation of TAE and OMPE in acetonitrile a

Scan Epa/V Epc/V ipa/PA ipclpA ipc/ipa ipav--ll2c--1/]2A rate/Vs --1 (vs. SCE) (vs. SCE) sl/2V--1/2m/~/-1

Tetra-p-anisylethylene (TAE): 2.19 mM 0.051 0.895 0.845 14.2 14.2 1.00 28.6 0.102 0.895 0.845 20.0 20.3 1.02 28.8 0.152 0.895 0.845 24.6 24.8 1.01 28.8 0.203 0.895 0.845 28.3 28.5 1.01 28.6 0.254 0.895 0.845 31.6 31.9 1.01 28.6 0.304 0.895 0.845 34.7 34.7 1.00 28.7 0.356 0.895 0.84 37.0 37.4 1.01 28.2 Tetrakis(p-N,N-dimethylaminophenyl)ethylene (OMPE): 1.23 0.051 0.163 0.131 9.5 9.7 1.0 0.102 0.161 0.131 13.4 13.7 1.02 0.152 0.161 0.131 16.7 17.0 1.02 0.203 0.161 0.131 18.9 19.1 1.01 0.254 0.161 0.131 21.2 21.4 1.01 0.304 0.16 0.13 23.7 24.1 1.02 0.356 0.16 0.13 26.2 26.9 1.01

mM 34 34.5 34.9 34.2 34.2 34.9 35.4

a The solution was 0.1 M in TBAP. The working electrode was a platinum disk with geo- metric area 0.013 cm 2.

330

separation between Epc and Epa is 30 mV; the separation between Epa ~ Epa/2 is 28.5 mV; the ratio of ipc/ipa is invariant with scan rate and equal to 1; the current function is invariant with scan rate and equal to about 34. The po- tential separations correspond very well with theoretical values of 29.5 mV and 28.2 mV expected for a simple, reversible two-electron oxidation [42]. Moreover, the current function is only slightly less than 23/2 times the value [13] found for a known one-electron oxidation of a similar compound, uTMPE. All of these data are consistent with a reversible two-electron oxidation of OMPE in ACN to a stable dication.

The cyclic vol tammetry of TAE in ACN is qualitatively the same as the ox- idation of OMPE in ACN: one undistorted reversible wave was observed. How- ever, quantitative results for TAE in ACN are quite different. The cyclic voltam- metric results (Table 5) show Epa --Epc of 50 mV and Epa --Epa/2 of 43 mV. These larger than theoretical potential separations cannot be ascribed to quasi- reversible charge transfer, since Epc , Epa, and the current function are essen- tially independent of scan rate. Nor are these larger potential separations at- tributable to uncompensated resistance, since the data for OMPE showed less than 1 mV deviation from theoretical values. These results are best explained by two successive, reversible one-electron transfers where the two E ° 's are very closely spaced. Using the working curve of Myers and Shain [43], one can cal- culate from Epa -- Epa/2 of 43 + 1 mV the separation between the E ° values (E ° - - E °) of the two steps as 7 + 4 mV.

The oxidation of OMPE in ACN at an RDE showed a single, well-defined wave with a half wave potential of 0.145 V vs. SCE and a Levich constant of 2.01. A log plot for this wave resulted in a straight line with slope -1 of 29.5 mV, in excellent agreement with the log plot slope -1 expected for a two-elec- tron oxidation. Thus both cyclic voltammetric and RDE results suggest that OMPE is oxidized in a direct two-electron transfer to a stable dication.

The oxidation of TAE at an RDE also showed a single, well-defined wave with an E~/2 of 0.875 V vs. SCE and a Levich constant of 1.97, about the same as that of OMPE.

The log plot for TAE in ACN, unlike that for MC, is quite linear. However, the slope -1 is 49 mV, which is intermediate between a slope -1 of 59 mV ex- pected for a one-electron oxidation and a slope -1 of 29.5 mV expected for a direct two-electron oxidation. Again, this apparent discrepancy can be recon- ciled by assuming that TAE in ACN is oxidized in the same stepwise mech- anism proposed for MC. The experimental log plot was fitted to the family of calculated curves for a stepwise mechanism; the best fit was obtained when K = 1.6, which corresponds to an E ° 12 + 5 mV more positive than E 0' in reason- able agreement with the cyclic voltammetric results.

Controlled potential coulometric oxidation of OMPE in ACN at 0.4 V vs. SCE required 2.0 electrons per molecule, corresponding to complete oxidation of OMPE to a dication. An e.s.r, experiment after complete electrolysis con- firmed the absence of a paramagnetic species. Reversal coulometry [38] also required 2.0 electrons per molecule; voltammetric results of the reduced dica-

331

tion were identical to the parent solution. Coulometric oxidation of TAE in ACN at 1.2 V vs. SCE required 2.1 electrons per molecule, again correspond- ing to the formation of a dication; an e.s.r, experiment also confirmed the absence of a cation radical. Reversal coulometry required slightly less than 2.0 electrons per molecule, and a RDE voltammogram of the reduced dication exhibited a smaller limiting current (some 5% less) than the original solution. Both observations indicate slow decomposit ion of the dication.

If the electrolysis was performed at .the foot of the wave, an nap p-value of less than 1.0 is obtained, and an e.s.r, spectrum shows the presence of a very intense singlet. Resolution of the hyperfine splittings which is attainable in methylene chloride [39] was never observed.

Electron spin resonance experiments on partially oxidized solutions

Although electrochemical data for the type (c) compounds, which show essentially a single two-electron oxidation to the dication, cannot provide in- formation about the separation between E ° and E °, direct observation of the radical cation in solutions containing both parent and dication can be used to estimate this separation. Thus for a type (c) compound the radical cation con- centration will vary during the electrolysis, reaching its maximum equilibrium concentration when the electrolysis has consumed one electron per molecule. To detect the radicals of type (c) compounds in MC, a solution of HMPE was partially oxidized, and an e.s.r, spectrum was obtained. The resulting spectrum was a moderately intense singlet similar to that obtained for TAE, thereby clearly demonstrating the existence of the radical cation in solution. Similar experiments with OMPE produced similar results. These results clearly dem- onstrate that the cation radical exists in significant concentrations in MC and further indicate that type (c) compounds probably undergo a stepwise oxida- tion with the second step occurring more easily than the first.

These results motivated some e.s.r, experiments in ACN. A solution of OMPE in ACN was partially oxidized and an e.s.r, experiment was conducted. The original experiment was completely negative; no signal was observed. How- ever, after accumulating the spectra with a Varian Model C-1024 Time Averag- ing Computer for 200 sweeps, the characteristic singlet was resolved. Thus OMPE is probably also oxidized in a stepwise manner in ACN. These results do not contradict the electrochemical results which indicate that OMPE is oxidized in a direct two-electron oxidation; rather, the results suggest that the second oxidation occurs at considerably more positive potentials than the first. In fact, from the e.s.r, and electrochemical results we can estimate that E ° in ACN lies in the range: (E ° -- 0.08 V) > E ° > (El -- 0.36 V).

DISCUSSION

The question of one- vs. two-electron oxidations has been dealt with pre- viously [1--3]. In general the second oxidation step involves removing an elec-

332

tron from a positively charged molecule where the electron repulsion of the second electron in the same molecular orbital is not present. From the gas phase ionization potentials the energy of removal of the second electron is of the order of 5.0 eV. In solution however, greater solvation of the dication and its greater tendency to form ion pairs decrease this separation; in the ab- sence of other factors this value is about 0.4 to 0.5 V for the usual aprotic solvents and supporting electrolytes. This behavior, characteristic of type (a) compounds, is observed, with somewhat closer spacing of the two oxidation steps, with uTMPE, and with TMPD.

In a more detailed consideration of the factors influencing the path and po- tential for the oxidation of substi tuted ethylenes the following factors must be considered:

Substituent effects

A comparison of the potentials for TMPE, HMPE and OMPE (Table 1) shows that the oxidation becomes easier with an increasing number of N,N-dimethyl (amino phenyl) groups, an expected result considering the strong electron donating characteristics of this group. Fritsch et al. [ 19] found similar trends with N,N-dimethylamino ethylenes. Furthermore, since the p-methoxy group is a less effective electron donating group than the p-N,N-methyl group, TAE is more difficult to oxidize than OMPE.

Solvent effects

The uncertainty in the junct ion potential values in the two solvents pre- cludes a direct comparison of the oxidation potentials of a given compound in the two solvents. However some information concerning possible solvent effects can be extracted from the difference in oxidation potential E ° -- E ° for a given compound in each of the solvents. For TAE E ° -- E ° was about 100 mV in MC and 12 mV in ACN, corresponding to equilibrium constants for the reproport ionat ion reaction (5) of 55 and 1.6, respectively. Similarly OMPE shows a larger concentrat ion of radical cation by e.s.r, upon partial oxidation in MC than in ACN. Thus the results indicate that the dication is more highly solvated in ACN, and that methylene chloride will show better resolution of stepwise electron transfers. Ammar and Sav6ant [2] have dis- cussed the thermodynamics of successive electron transfers and solvation enthalpy and entropy effects for the stepwise reduction of polynitro-com- pounds in DMF and ACN.

Steric effects

A surprising aspect of the oxidation of the type (c) compounds is the ease ~)f the second oxidation, so that the second oxidation occurs at potentials equal to or more negative than the first. A plausible explanation is that con-

333

formation changes in the ethylenes occur as the oxidation state changes. Mod- els show that tetraphenylethylene-related compounds exhibit considerable steric strain when constrained to planarity because of the cis o-hydrogen in- teractions. To relieve the strain, either rotat ion about the C-phenyl bond or torsion about the ethylene double bond occurs. Garst et al. [44] have invoked a similar qualitative explanation to account for the disproport ionation of the tetraphenylethylene anion radical. To examine these effects, discussed in more detail in ref. 1, Goldberg [45] performed HMO calculations incorporating varying degrees of rotat ion and torsion. Although the interpretation is nec- essarily only semi-quantitative, these calculations suggest that the bond order of the C-phenyl bond increases and that the bond order of the ethylene double bond decreases as the molecule goes from the parent to the cation radical and then to the dication. Thus as the ethylene undergoes oxidation to the cation radical, torsion about the double bond can occur, thus elevating the relative position of the highest occupied molecular orbital and facilitating the removal of the second electron. These conformational changes were experimentally supported by the reversal coulometric experiments on cis-sTMPE which clearly demonstrated that isomerization of the cation radical, of the dication, or of both occurred in solution. Similar isomerizations of radical anions are ob- served in the reductions of substi tuted ethylenes such as stilbene and diethyl maleate, where addition of an electron to the antibonding orbital weakens the ethylenic double bond.

Finally the results found here are relevant to past studies of similar systems. Kuwata and Geske [16] showed that te tra(dimethylamino)ethylene undergoes two stepwise one-electron oxidations in ACN. Fritsch et al. [18,19] extended this work to include a number of other dimethylamino-substi tuted ethylenes. In this s tudy too steric repulsion be tween C- and N-methyl groups can lead to cyclic voltammetric waves characteristic of direct two-electron oxidations. While the original s tudy of Parker et al. [15] suggested a two-electron oxida- tion of TAE in ACN + LiCIO4, the results presented here, as well as a recent s tudy of the Parker group [47], show that the reaction occurs by stepwise one-electron transfers in both MC and ACN (where in ACN E ° is about 7 mV more positive than E °) and points to the importance of solvation in the separa- tion between the E ° values.

Two other studies pert inent to the oxidation of TAE have appeared. A series of investigations of the complexes of TAE with various oxidants has culminated in the reported crystal structure of a salt Of the dication of TAE [40]. Valenzuela and Bard [39] have reported an e.s.r, s tudy of the cation radical of TAE. All these are thus consistent with stepwise oxidation of TAE.

We have shown that even with the type (c) compounds such as OMPE, which show two-electron electrochemical behavior, appreciable quantities of radical cation can be produced. These results can explain some previous chemical studies [41,46]. Anderson et al. [41] noted that the diiodide salt of OMPE dication exhibited no e.s.r, activity when dissolved in water, whereas this same salt exhibited considerable e.s.r, activity when dissolved in ethylene ch lo r ide .

334

In fact, their results indicated some 20% of the compound existed as the cation radical in e thylene chloride. To account for this curious behavior, they postu- lated, in somewhat modif ied form, the following equilibria:

OMPE 2+ + I - ~ OMPE +" + 1/2I 2 (9)

OMPE +" + I - ~ OMPE + 1/2I 2 (10)

They fur ther postulated that the difference noted could be at tr ibuted to sol- vent effects: more polar solvents favor more highly charged species and less polar solvents favor the less highly charged species. Elofson et al. [46] have also invoked, on the basis of e.s.r, and n.m.r , studies, the intermediacy of OMPE cation radical in the autoxidat ion of OMPE in benzene.

The equilibrium constants for (9) and (10) can be calculated from the stan- dard potentials of the relevant half reactions in MC. E N (OMPE2÷/OMPE) = (E ° + E ° ) / 2 = 0.16 V and E°(I2/I - ) = 0.20 V (estimated from cyclic voltam- metric measurements of a 0.2 M TBAI + MC solution using a plat inum elec- trode). Assuming different values of AE ° = E ° -- E ° , for a 1 mM solution of OMPE, standard equilibrium calculations show that 3 to 6% of the OMPE will be in the form of the cation radical for AE ° values of 60 to 0 mV. Thus a solution of (OMPE 2÷) (I- )2 should exhibit e.s.r, activity in MC and probably also in ethylene chloride.

ACKNOWLEDGMENTS

The support of the Rober t A. Welch Foundat ion (F-079) and the National Science Foundat ion (GP-31414X) is gratefully acknowledged.

REFERENCES

1 A.J. Bard, Pure Appl. Chem., 25 (1971) 379, and references therein. 2 F. Ammar and J.M. Say,ant, J. Electroanal. Chem., 47 (1973) 115, and references

therein. 3 A.J. Bard and J. Phelps, J. Electroanal. Chem., 25 (1970) App. 2. 4 H. Schmidt and H. Meinert, Angew. Chem., 72 (1960) 109. 5 F. Mango and W. Bonner, J. Org. Chem., 29 (1964) 1367. 6 L. Eberson and K. Nyberg, J. Amer. Chem. Soc., 88 (1966) 1686. 7 L. Eberson and K. Nyberg, Acta Chem. Scand., 18 (1964) 1568. 8 T. Inove, K. Koyama, T. Matsuoka, K. Matsuoka and S. Tsutsumi, Tetrahedron Lett.,

(1963) 1409. 9 T. Inove and S. Tsutsumi, Bull. Chem. Soc. Jap., 38 (1965) 661.

10 J.J. O'Conner and I.A. Pearl, J. Electrochem. Soc., 111 (1964) 335. 11 V.D. Parker and L. Eberson, Chem. Commun., (1969) 340. 12 H. Sch{ifer and E. Steckhan, Angew. Chem., Int. Ed., 8 (1969) 518; H. Sch~ifer, Chem.

-Ing.-Techn., 42 (1970) 164; H. Sch~ifer and E. Steckhan, ibid., 44 (1972) 186. 13 J.D. Stuart and W.E. Ohnesorge, J. Amer. Chem. Soc., 93 (1971) 4531. 14 J.D. Stuart and W.E. Ohnesorge, J. Electroanal. Chem., 30 (1971) App. 15 V.D. Parker, K. Nyberg and L. Eberson, J. Electroanal. Chem., 22 (1969) 150.

335

16 K. Kuwata and D. Geske, J. Amer. Chem. Soc., 86 (1964) 2101. 17 S. Hi~nig, H. Schlaf, G. Kiesslich and D. Scheutzow, Tetrahedron Lett. , (1969) 2271;

S. Hiinig and J. Gross, ibid., (1968) 2599; S. Hiinig, Angew. Chem., Int. Ed., 8 (1969) 286.

18 J.M. Fritsch, H. Weingarten and J.D. Wilson, J. Amer. Chem. Soc., 90 (1968) 793. 19 J.M. Fritsch, H. Weingarten and J.D. Wilson, J. Amer. Chem. Soc., 92 (1970) 4038. 20 N. Wiberg and J.W. Buchler, Angew. Chem., 74 (1962) 490; N. Wiberg and J.W. Buchler,

Ber., 96 (1963) 3223; N. Wiberg, Angew. Chem., Int. Ed., 7 (1968) 766. 21 R.D. Hoffman, Angew. Chem., Int. Ed., 7 (1968) 754. 22 C.K. Mann in A.J. Bard (Ed.), Electroanalytical Chemistry, Vol. 3, Marcel Dekker,

New York, 1969, Ch. 2. 23 D.F. Shriver and J. Posner, J. Amer. Chem. Soc., 88 (1966) 1672. 24 L. Gatterman, Ber., 28 (1895) 2869. 25 R.E. Buckles and N.A. Meinhardt, J. Amer. Chem. Soc., 74 (1952) 1171. 26 J.E. Phelps, Ph.D. Dissertation, The University of Texas at Austin, 1972. 27 F.C. Anson and D.A. Payne, J. Electroanal. Chem., 13 (1967) 35. 28 A.J. Bard and E. Solon, J. Phys. Chem., 67 (1963) 2326. 29 A.J. Bard and K.S.V. Santhanam, J. Amer. Chem. Soc., 88 (1966) 2669. 30 V. Dvorak, I. N~mec and J. Z:~ka, Microchem. J., 12 (1967) 324. 31 D.S. Polcyn and I. Shain, Anal. Chem., 38 (1966) 370. 32 I. Shain and R. Wopschall, Anal. Chem., 39 (1967) 1514. 33 E. Laviron, J. Electroanal. Chem., 52 (1974) 355, and references therein. 34 O.H. Miiller, Ann. N.Y. Acad. Sci., 40 (1940) 91. 35 R. Brdi~ka, Z. Elektrochem., 47 (1941) 314. 36 J.H. Christie, R.A. Osteryoung and F.C. Anson, J. Electroanal. Chem., 13 (1967) 236. 37 F.C. Anson, J.H. Christie and R.A. Osteryoung, J. Electroanal. Chem., 13 (1967) 343. 38 A.J. Bard and S.V. Tatwawadi, J. Phys. Chem., 68 (1964) 2876. 39 J.A. Valenzuela and A.J. Bard, J. Phys. Chem., 72 (1968) 286. 40 N.C. Baenziger, R.E. Buckles and T,D. Simpson, J. Amer. Chem. Soc., 89 (1967) 3405. 41 D.H. Anderson, R.M. Elofson, H.S. Gutowsky, S. Levine and R.B. Sandlin, J. Amer.

Chem. Soc., 83 (1961) 3157; R.M. Elofson, D.H. Anderson, H.S. Gutowsky, R.S. Sandlin and K.F. Schulz, J. Amer. Chem. Soc., 85 (1963) 2622.

42 R.S. Nicholson and I. Shain, Anal. Chem., 36 (1964) 706. 43 R.L. Myers and I. Shain, Anal. Chem., 41 (1969) 980. 44 J.F. Garst, E.R. Zabolotny and R.S. Cole, J. Amer. Chem. Soc., 86 (1964) 2259; J.F.

Garst, J.G. Pacifici and E.R. Zabolotny, ibid., 88 (1966) 3872; J.F. Garst, ibid., 93 (1971)6312.

45 I. Goldberg, unpublished results, 1970. 46 R.M. Elofson, K.F. Schulz and R.B. Sandlin, Cand. J. Chem., 47 (1969) 4447. 47 U. Svanholm, B.S. Jensen and V.D. Parker, J .C.S.P.T., II (1974) 907.